Hello, I am trying to build tensorflow from source on a small Rocks cluster. Unfortunately, when I use my own tensorflow build I get only 30 % GPU utilization. For testing purposes I use cifar10 dataset, which was cloned using git clone command. On the other hand, when I am trying to execute the same cifar10_train.py script inside a tensorflow docker container I get over 60 % of GPU utilization. May someone give me some hints how can I boost GPU utilization or where does docker saves information about how docker container was built (in my case nvcr.io/nvidia/tensorflow:19.04-py2)
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Tensorflow:17.10 image doesn't work out of the box | 6 | 3055 | October 12, 2021 | |
| Docker container cant use GPU | 1 | 4446 | July 1, 2022 | |
| Full processing only in one GPU, how to leverage the processing to another GPU? | 1 | 1038 | April 16, 2019 | |
| Docker running Time | 0 | 585 | February 24, 2019 | |
| Tensorflow can not find gpu device | 2 | 590 | March 23, 2021 | |
| Access to the Dockerfile from which the image was built. | 0 | 576 | November 27, 2019 | |
| Tensorflow docker can't detect gpu | 0 | 3911 | December 31, 2020 | |
| Linux Server setup suggestions for remote CUDA and Tensorflow. | 2 | 1220 | March 28, 2017 | |
| When running containers all gpus are visible in contaner regardless of --gpus option settings | 0 | 1319 | November 23, 2021 | |
| Tensorflow and CUDA 12 | 6 | 42768 | January 3, 2023 |